Page History
...
- observation errors are represented by perturbing observations. The observations are assumed unbiased (once any dynamic bias correction has been applied) and observation errors are assumed to have a normal distribution.
- model errors are represented by perturbing the forecast model. Model error is simulated using the Stochastically Perturbed Parameterisation scheme (SPP).
- boundary condition errors are represented by perturbing soil moisture, sea-surface temperature, and sea ice across appropriate length scales.
...
- Read more about quantifying forecast uncertainty.
- Read more on the ensemble of data assimilations EDA at ECMWF.
- Read more about Stochastically Perturbed Parametrization (SPP) scheme.
(FUG Associated with Cy49r1)
...
Overview
Community Forums
Content Tools